Hi,
I'm trying to figure out the best way to do my higher level analysis.
Our dataset consists of 21 subjects with 4 runs each of the same paradigm. I
ran feat on each run(84 runs). I'm trying to see between which 2 runs we
find significant difference in activation(in other words inter-run
variability for eg due to physiological reasons like habituation)or if
theres a trend in the level of activation in the 4 runs. To achieve the
former i found single group mean for each run and then checked different
pairs for significant differences using 3rd level analysis. Is this correct?
i'm not so sure how to check for a trend in the 4 runs.
Also,would it matter whether i used Fixed effects or Mixed effects analysis
for 2nd level and/or 3rd level analysis?
Thanks!
cheers,
Vish
|